SimRelUz

SimRelUz是由乌尔根奇州立大学和科鲁尼亚大学合作创建的乌兹别克语语义评估数据集，包含1418对精心挑选的词汇，涵盖多种词性、词形和频率，以及罕见的和超出词汇表的词汇。数据集通过11位不同背景的乌兹别克语母语者的标注，旨在评估语义模型在处理乌兹别克语时的性能。该数据集适用于自然语言处理中的多种应用，如词义消歧、自动词典生成和机器翻译等，以解决低资源语言在语义分析方面的挑战。

SimRelUz is an Uzbek semantic evaluation dataset jointly developed by Urgench State University and the University of A Coruña. It contains 1,418 carefully selected word pairs, covering various part-of-speech categories, word forms, frequency levels, as well as rare and out-of-vocabulary (OOV) words. Annotated by 11 native Uzbek speakers with diverse backgrounds, this dataset is designed to evaluate the performance of semantic models when processing the Uzbek language. It is applicable to multiple natural language processing (NLP) applications such as word sense disambiguation, automatic dictionary generation and machine translation, aiming to address the challenges in semantic analysis for low-resource languages.

搜集汇总

背景与挑战

背景概述

SimRelUz是由乌尔根奇州立大学和科鲁尼亚大学合作创建的乌兹别克语语义评估数据集，包含1418对词汇，涵盖多种词性、词形和频率，包括罕见词汇，通过11位母语者标注，旨在评估语义模型在乌兹别克语中的性能。该数据集适用于自然语言处理中的词义消歧、自动词典生成和机器翻译等应用，以解决低资源语言在语义分析方面的挑战。